• 한국문헌정보학회지
• /
• 제39권2호
• /
• pp.123-146
• /
• 2005

이 연구에서는 문서 자동분류에서 분류자질 선정과 가중치 할당을 위해서 일관된 전략을 채택하여 kNN 분류기의 성능을 향상시킬 수 있는 방안을 모색하였다. 문서 자동 분류에서 분류자질 선정 방식과 자질 가중치 할당 방식은 자동분류 알고리즘과 함께 분류성능을 좌우하는 중요한 요소이다. 기존 연구에서는 이 두 방식을 결정할 때 상반된 전략을 사용해왔다. 이 연구에서는 색인파일 저장공간과 실행시간에 따른 분류성능을 기준으로 분류자질 선정 결과를 평가해서 기존 연구와 다른 결과를 얻었다. 상호정보량과 같은 저빈도 자질 선호 기준이나 심지어는 역문헌빈도를 이용해서 분류 자질을 선정하는 것이 kNN 분류기의 분류 효과와 효율 면에서 바람직한 것으로 나타났다. 자질 선정기준으로 저빈도 자질 선호 척도를 자질 선정 및 자질 가중치 할당에 일관되게 이용한 결과 분류성능의 저하 없이 kNN 분류기의 처리 속도를 약 3배에서 5배정도 향상시킬 수 있었다.

• 한국해양공학회:학술대회논문집
• /
• 한국해양공학회 2004년도 학술대회지
• /
• pp.85-89
• /
• 2004

The paper presents a summary of the multidisciplinary/optimization method for the preliminary design of container feeder vessel. The current scenario in the ship building industry highly focuses on container ship design and construction proving the inherent demand in maritime industry. The design accomplishes the outer circle of the design spiral giving stress in areas of Hull Form Design, Resistance & Propulsion. Empirical relations, model test results, data from built ships, class rules and latest market demands stood as the criteria for the design. Optimization of the design as per the owners requirement. class rules, and the trade route selected are the major challenges met with. Strength. reliability, structural safety and stability have been incorporated in compromising standards.

• Communications for Statistical Applications and Methods
• /
• 제18권3호
• /
• pp.267-275
• /
• 2011

According to recent studies, Bayesian information criteria(BIC) is proposed to determine the structural dimension of the central subspace through sliced inverse regression(SIR) with high-dimensional predictors. The BIC may be useful in K-means clustering inverse regression(KIR) with high-dimensional predictors. However, the direct application of the BIC to KIR may be problematic, because the slicing scheme in SIR is not the same as that of KIR. In this paper, we present empirical penalty term studies of BIC in KIR to identify the most appropriate one. Numerical studies and real data analysis are presented.

• 한국주거학회논문집
• /
• 제13권3호
• /
• pp.103-113
• /
• 2002

The purpose of this study is to find out the characteristics of users'behavior for the planning of the shared community space(SCS). It was analyzed and estimated that the present status of the SCS of the contemporary multi-family housing sites in Korea and the residents'behavior in the SCS. Theoretic investigation and empirical research methods were used in this study. The empirical survey focused on seeking out user-oriented design criteria, based on the analysis of residents'usage behavior for the SCS. The analysis was made by both quantitative and qualitative methods. The results of this study are as follows; First, as to the usage status survey for the SCS, there is a difference in the degree of usage according to their locations. Second, as to the analysis of the SCS through a mental map, the sketch map analysis is found to be a useful research method for the SCS planning by actualizing the residents'behavior characteristics. In detail, the main characteristics in the SCS planning are nearness, centrality, intimacy, combination of facilities and the connection with the main paths of the complex.

• Journal of Astronomy and Space Sciences
• /
• 제32권2호
• /
• pp.137-140
• /
• 2015

The results of the study of ionospheric variations in the summer months of 1998-2002 at an ionospheric station of vertical sounding "Petropavlovsk-Kamchatsky" are presented. Anomalous variations of virtual sporadic-E height (h'Es), Es blanketing frequency (fbEs), and the critical frequency of the ionospheric F2 layer (foF2) (which can be attributed to the possible earthquake precursors) are selected. The high efficiency of the selection of ionospheric earthquake precursors based on the several parameters of Es and F2 layers is shown. The empirical dependence, which reflects the connection between the lead-time of the earthquake moment, the distance to the epicenter from the observation point, and the magnitude of the earthquake are obtained. This empirical dependence is consistent with the results of the detection of earthquake precursors by measuring the physical parameters of the Earth's crust in the same region.

• 한국전자거래학회지
• /
• 제6권2호
• /
• pp.143-157
• /
• 2001

There has been a need for predicting development efforts and costs of the system during the early stage of the software process and hundreds of metrics have been proposed for computer software, but not all provide practical support to the software engineer. Some demand measurement that is too complex, others are so esoteric that few real-world professionals have any hope of understanding them, and others violate the basic intuitive notions of what high-quality software really is. It is worthwhile that metrics should be tailored to best accommodate specific products and processes after grasping their good and no good point. This paper describes two size estimation techniques, the Karner technique and the Marchesi technique, and compares and analyzes them with proposed evaluation criteria. Both techniques are to estimate software size analyzed by use case that is mainly described during the object-oriented analysis phase. We also present an empirical comparison of them, both are applied in the Internet Medicine Prescription System. We also propose some guidance for experiments based on our analysis. We believe that it should be facilitating project management more effective by adjusting software metrics properly.

• 한국의류산업학회지
• /
• 제12권3호
• /
• pp.318-326
• /
• 2010

The advantages of the multi-channel retailing have been widely discussed but empirical research on fashion multi-channel retailing has been limited. In this study, multi-channel concept was discussed and then, channel choosing condition of apparel shopper and channel choosing criteria for information search and buying were investigated as a empirical study. Drawing on a sample of 298 customers of apparel products in Korea, the result demonstrated that some differences in the perception of experience goods and search goods among apparel products. And, according to buying channel, consumers were different from each other in information search time and clothing expenses. Some suggestion for the future research of multi-channel retailing was given.

• 품질경영학회지
• /
• 제27권1호
• /
• pp.195-210
• /
• 1999

The multiplicity of software packages brings about a choice problem to the end-user and a multi-fold intensity to the competing developer. This study intends to provide guidelines solving the problems for both the end-user and the developer of packages for automatic statistical analyses. For the purpose, we adopt an appropriate evaluation model in terms of user-perceived quality through the applicable criteria and an evaluation method. The result of empirical study is described, which shows that the weight of each criterion is dynamic along usage levels. Briefly, functionality, usability, and reliability are key factors to contribute to the perceived quality. The empirical study also illustrates a comparative analysis on the dominant products, SAS and SPSS, in Korea.

• Ocean and Polar Research
• /
• 제44권4호
• /
• pp.339-353
• /
• 2022

The purpose of this research is to conduct an empirical analysis regarding Korea's 'Marine Protected Areas (MPA)' from an international comparative perspective. The authors would like to present the policy direction of marine protected areas based on the result of the research. The result of this study can be summarized in the following manner: First, the total designated size of marine protected areas in Korea is very small compared to the sizes of other OECD nations . Second, while some nations have expanded the extent of marine protected areas in accordance with international agreements and criteria, Korea has not done so. Accordingly, we propose the designated dimensions of marine protected areas should be constantly expanded to keep pace with international trends.

• 정보처리학회논문지B
• /
• 제13B권7호
• /
• pp.633-642
• /
• 2006

현재 인터넷 환경에서 사용자는 서로 잘 모르는 사람이나 시스템과 상호거래를 하게 되는데 이 경우 서로 다른 개체에 대한 신뢰 정보가 부족하기 때문에 상호 거래의 위험을 감수할 수밖에 없다. 따라서 이러한 불확실성과 위험을 감소시킬 수 있는 방안으로 상대 개체와 직접 경험한 신뢰정보와 추천자에 의한 명성정보를 계산하여 이를 활용하는 방법들이 대두되고 있다. 이 논문에서는 개체에 대한 신뢰를 계산하기 위해 상호거래 결과를 누적한 경험적 확률분포와 여러 가지의 평가 기준에 의한 만족도를 계산하고, 이를 다른 개체들로부터의 추천정보와 결합하여 계산하는 신뢰 모델을 제안한다. 제안한 모델에서는 개체의 신뢰도를 개체가 주어진 상황에서 만족스러운 결과를 낼 기대값으로 정의하고, 다른 개체와 상호작용이 일어날 때마다 각 평가 기준에 따른 평가결과가 얻어진다고 전제한다. 신뢰 정보가 요구될 때 우선 경험적 확률분포와 개체의 평가결과에 대한 선호도를 고려하여 각 평가 기준에 대한 만족도를 계산하고, 계산된 만족도 값들은 각 평가기준의 중요도를 반영하여 하나의 신뢰값으로 결합되며, 이때 추천 정보도 신뢰값에 함께 결합되는 모델이다. 이 논문에서는 제안한 모델을 이용해 전자상거래에 적용한 실험 결과를 보여 주고 있다.